Metabolomics in Natural Product Analysis

Tea is one of the most widely consumed beverages in the world and needs to be protected from changing climate conditions as much as possible. Amanda Kowalsick and Albert Robbat of the Department of Chemistry at Tufts University, Massachusetts, USA, are working to understand how vulnerable the metabolites in tea are to climate conditions. They recently spoke to LCGC about their work with tea and natural products. Q. What role does metabolomics play in the analysis of plant samples?

A: Plants produce a diverse array of secondary metabolites that assist in their protection and reproduction. Metabolites in plant products we eat and drink affect quality including flavour, aroma, and health benefits. Chemical composition is species-dependent and influenced by both human, (for example, farming and processing) and environmental factors (such as changes in climate and soil).

Although gas chromatography coupled to mass spectrometry (GC–MS) can provide a great deal of information, natural products are so complex that GC–MS can miss more than 50% of compounds in the sample. Moreover, our senses can detect analytes at lower levels than current instrumentation. Since many important flavour and nutraceutical compounds are masked by the sample, two-dimensional (2D) GC–MS can provide the separation space needed to identify these constituents. Once identified, we can use spectral deconvolution of one-dimensional (1D) GC–MS data to produce metabolomic profiles in food and beverage products.


Q. How has metabolomics profiling of natural products evolved, and what are the most recent trends?

A: 2D GC–MS is only limited by the sample preparation needed to produce a representative extract of the sample and by the time it takes to completely profile that sample. Over the past decade many new sample introduction systems have been developed, from solid-phase extraction (SPE), solid-phase microextraction (SPME), and stir-bar sorptive extraction (SBSE) techniques to total evaporation of the sample in the instrument. These automated sample introduction systems produce more comprehensive profiles in less time.


Q. What are the benefits of using 2D GC–MS for complex natural food products?

A:  We use 2D GC–MS (both GC–GC–MS and GCxGC–MS) to obtain libraries of retention time and mass spectral data of compounds in the sample. Once we have this we use spectral deconvolution to analyze the data by 1D GC–MS. 2D GC–MS increases the separation space compared to GC–MS in a manner that makes it easy for the analyst to confirm compound identity. The combination of these technologies offers several distinct advantages over conventional approaches, including the ability to obtain unencumbered retention times and mass spectra of metabolites free of matrix interferences.


Q. You recently studied the effect of seasonal changes on metabolite profiling of tea. Could you tell us more about this research?

A: In this study, tea (Camellia sinensis (L.) Kuntze; Theaceae) served as a model system documenting the role extreme weather events might have on plant physiology. Data show extreme rainfall events associated with the East Asian Monsoon are starting earlier and lasting longer. As a result, the harvest window to obtain high quality tea from the spring season is narrowing. Our collaborators found that catechins and methylxanthines, responsible for bitter, astringent notes in high quality Yunnan tea, decreased by 50% from the spring to monsoon season (1).

In our study, we identified ~ 200 metabolites in the tea (2). Of these, 169 were common to both seasons. We also found ~ 30 compounds unique to each season. Striking concentration differences were observed in as little as five days after the monsoon onset, with individual metabolites in each chemical family increasing, decreasing, or remaining the same. Spring teas had higher concentrations of compounds known to impart floral, sweet, and honey-like characteristics. Our results were consistent with the sensory experience of farmers in the region: namely, spring teas contained organics usually associated with sweet, floral aroma compared to the monsoon tea, which exhibited green, earthy flavours. Our colleagues will soon conduct a consumer-facing study to determine how tea quality (flavour and nutraceuticals) affects purchasing decisions (3).


Q. Food and beverage companies now need to be able to track the genealogy of their products - from the raw materials to the final product. Could you talk a little about this in the light of your research with juniper berries and gin?A: One way of tracking product quality from farm to final product is to identify the key biomarkers that survive environmental, farming, and manufacturing processes. These biomarkers should also be those responsible for flavour and health benefits ascribed to products by the manufacturers. Metabolomic profiling is the key to finding these indicators of quality and authenticity. Spectral deconvolution of the data allows us to track these compounds even when the final product becomes more complex by the addition of other botanicalsand essential oils, or both.  For example, gin typically contains 6–12 essential oils and botanicals. Some compounds are common to all ingredients; others are not. It is these unique biomarkers that differentiate one product from another that allow us to authenticate the product manufacturer, and also allows us, through spectral subtraction, to find pollutants and adulterants.


Q. Where will your research take you in the future?A: Our work demonstrates that 2D GC–MS is a powerful analytical technique capable of producing matrix-specific libraries of complex natural products. Spectral deconvolution of 1D GC–MS data based on these libraries provides a reliable, unambiguous means to profile the metabolites in natural products to determine product and manufacturing quality, and assess seasonal variation in plant materials chemistry and quality. Soon we will begin to do the same with liquid chromatography coupled to MS (LC–MS). We will use LC–LC–MS to build libraries and LCxLC–MS to quantify analytes in the samples. Spectral deconvolution of the data is central to our investigation; in the case of LC, we will increase the ionization voltage to obtain the spectral information we need. The combination of 2D GC and LC information should provide the data we need to unravel how plants respond to changing environmental conditions. We are interested in learning, for example, what are key constituents in tea, independent of location.


References

1. S. Ahmed, J.R. Stepp, C. Orians, et al., PloS one9, e109126 (2014).

2. A. Kowalsick, N. Kfoury, A. Robbat Jr, et al., Journal of Chromatography A1370, 230 (2014).

3. Boehm et al., "Reading the tea leaves of climate change: consumer preference for seasonal variability in brewed green tea and results from a choice experiment on willingness to pay for tea product attributes,” paper to be presented at the Agricultural & Applied Economics Association Conference, July 2015.

Amanda Kowalsick is a postdoctoral researcher in the Department of Chemistry at Tufts University, Massachusetts, USA. Dr. Kowalsick’s research interests include the development of analytical tools and methods such as automated sequential two-dimensional gas chromatography–mass spectrometry. This research has led to total metabolomic profiles of plant-based materials such as botanicals and their essential oils, gin, and tea. These studies have resulted in new methods to authenticate products, establish new quality control standards, and to differentiate seasonal differences in tea.
 

Albert Robbat is a professor in the Department of Chemistry, Tufts University. Dr. Robbat’s research interests include the development of innovative analytical instruments, methods, and data analysis software used to solve a wide range of environmental, food, and water problems. His research includes forensic investigations of environmental pollutants, remediation of heavy hydrocarbon hazardous waste sites, elimination of fat, oil, and grease in wastewater collection and treatment systems, and beach restoration using different combinations of plant extracts as well as metabolomic profiling of plant materials.